Cytotoxic antitumor agents usually cause severe side-effects because of adverse effects on normal tissue cells. The long-term objective of this project is the development of such a strategy for the amplification of tumor-to-host specificity through the use of self-assembling cytotoxins. These molecules will self-assemble in vivo through covalent bond formation between smaller, less toxic precursor molecules. Because the velocity of second order reactions is proportional to the product of the reactant concentrations, any selection tumor-to-host affinity exhibited by the individual reactant molecules will be amplified in the combination. THE IMPROVEMENT OF EXISTING TUMOR- SPECIFIC MOLECULES THROUGH SPECIFICITY AMPLIFICATION INVOLVING SELF-ASSEMBLING CYTOTOXINS WOULD LEAD DIRECTLY TO AN ENTIRE SERIES OF POWERFUL NEW ANTITUMOR PROTOCOLS. The specific aims of this project are as follows: (1) Demonstration of specificity amplification using erythrocytes as models for tumor cells. The self-assembling cytotoxin will consist of two components, such as a hydrazine derivative and an aldehyde which combine to produce a more cytolytic hydrazone. The dose-response of untreated and neuraminidase-treated erythrocytes lysis to the combination will be compared to the responses to the individual components and the preformed product in order to ascertain that the self-assembling combination is the more selectively cytotoxic against the more anionic, untreated erythrocytes. (2) Demonstration of antineoplastic specificity amplification using rhodamine-derived cationic aldehydes and hydrazine derivatives which self-assemble into simple monoadducts. Adduct toxicity will derive from membrane disruptive and/or metal chelating abilities greater than those of the precursors. (3) Self-assembling antineoplastic polymer formation. Combinations of bis-hydrazine derivatives and bis-aldehyde derivatives of cationic rhodamine dyes will be tested for selective cytotoxic activity against cultured carcinoma as opposed to normal cell lines.